Emmanuel Lecharny wrote:
Hi Alex, hi Stefan
<snip/>
Right and there will also be character encoding issues. Hmmm I think this switch from using a String to using a byte[] should be based on the attributeType definition. Question is how do we get that to the Rule in a safe and easy manner?I've tested an AddRequest with a PDU which contains a jpegPhoto and a byte[] as its value, and I'm not sure that the org.apache.ldap.common.berlib.asn1.decoder.add.AddAttributeValueRule.finish() method does the right thing when creating a String with the bytes it get. the line before the last line is : ... // Add the attribute attrs.put( id, new String( octets ) ) ; ...and should be ...// Add the value attrs.put( id, octets ) ; ... If the bytes are above 128, you will just transform the bytes to a bad Value. I think that the value, as it is supposed to be an OctetString, should be stored directly without creating a new String.wdyt Alex ?
Alex
